flowpoint commited on
Commit
2121e0b
·
1 Parent(s): cfd2a25

add fasttext model, fasttext binary and readme

Browse files
Files changed (5) hide show
  1. .gitattributes +3 -0
  2. README.md +38 -0
  3. fasttext +3 -0
  4. model1.bin +3 -0
  5. model1.vec +3 -0
.gitattributes CHANGED
@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ model1.vec filter=lfs diff=lfs merge=lfs -text
37
+ model1.bin filter=lfs diff=lfs merge=lfs -text
38
+ fasttext filter=lfs diff=lfs merge=lfs -text
README.md CHANGED
@@ -1,3 +1,41 @@
1
  ---
2
  license: apache-2.0
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
  ---
4
+
5
+ # fasttext cbow on dclm400
6
+
7
+ A continuous-bag-of-words model trained on https://huggingface.co/datasets/mlfoundations/dclm-pool-400m-1x
8
+
9
+ the cbow model was trained with https://github.com/facebookresearch/fastText/
10
+
11
+ the dataset was downloaded with git-lfs
12
+
13
+ the dataset commit was: f20ae752116ce7b4ab15d31e1e40b094229bf911
14
+
15
+ the files decompressed with:
16
+
17
+ `parallel "zstd --keep --stdout -d {} | jq .text > {/}.txt" ::: /root/lfs/dclm-pool-400m-1x/*.jsonl.zst`
18
+
19
+ concatenated with
20
+
21
+ `cat *.txt > CC_SHARD_ALL.jsonl.txt`
22
+
23
+ the `sha256sum CC_SHARD_ALL.json.txt` is
24
+
25
+ `576e4e79e76b9ca24dc77a8da0df17ad5efc9c5ca16c9a86f62e7b7b4ae8c640 CC_SHARD_ALL.jsonl.txt`
26
+
27
+ then the fasttext model was trained with defaults settings from
28
+
29
+ compiled with gcc 13.3.1
30
+
31
+ fasttext-repo (main branch) with the commit hash `1142dc4c4ecbc19cc16eee5cdd28472e689267e6`
32
+
33
+ training command:
34
+
35
+ `prlimit -m 3200000000 fasttext cbow -input CC_SHARD_ALL.jsonl.txt -output fasttext_models/model`
36
+
37
+ the exact fasttext binary is included in this repo as `fasttext`
38
+
39
+ the decompression and concatenating took a few hours.
40
+
41
+ the model training took 100 hours on 8 cores plus a few hours to read in the words (fasttext)
fasttext ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2eb67f89e29c7da08b3cdbc4b1e9219bb6c564a3c57f399270ea524a23241a43
3
+ size 473848
model1.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e6c28036e8feb26266f749e798cf3262f647b4a96ff351e37a0e3cd171970cf
3
+ size 3588749523
model1.vec ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd0a30892361f8581c04bb32bbfb4e212a1384631be70585d0a64bac3c4d9028
3
+ size 3916574498